AN ABSTRACT OF THE THESIS OF

RICHARD WALTZ HARPER for the MASTER OF SCIENCE (Name) (Degree)

in CHEMISTRY presented on ()CVON° l(-0\\

(Major) (Date)

Title: A STUDY OF CERTAIN NOVEL REACTIONS OF 4, 6-

DIME T HOX Y -NI T ROP YRIMID INE

Abstract approved: B E hristénsen

Reaction of 4, 6- dimethoxy -5- nitropyrimidine [I], with reflux-

ing pyridine, was discovered to yield the methylpyridinium salt of

1, 6- dihydro -4- hydroxy -1 -methyl -5 -nit ro -6- oxopyrimidine [III]. Pos-

sible use of I as a general N- methylating agent was explored. No

appreciable reaction occurred between I and refluxing excess n -butyl

alcohol. Similar reaction involving quinoline failed to yield an isol-

able quinolinium salt although the reaction gave an oily product.

Equimolar amounts of I and pyridine in refluxing ethanol solvent gave

no evidence of a reaction. As a means of minimizing isolation prob-

lems, equimolar amounts of I and pyridine without solvent were re-

acted. Under these conditions a second product resulting from rear-

rangement of I, 1, 6-dihydro-4- methoxy -l- methyl -5- nitro -6-

oxopyrimidine [V] was isolated in addition to III. Fusion of I with

Ct(-3

equimolar amounts of piperidine, quinoline, and 4- methylpyrimidine,

resulted in smaller amounts of V, and a black foamy solid, with no

detectable methylation of the solvent.

s

A Study of Certain Novel Reactions of 4, 6- dimethoxy -5 -nit ropyrimidine

by

Richard Waltz Harper

A THESIS

submitted to

Oregon State University

in partial fulfillment of the requirements for the

degree of

Master of Science

June 1968

APPROVED:

Professor of Chemistry

in charge of major

Chairman of Department of Chemistry

Dean of Graduate School

Date thesis is presented D L16b`C t ) `-1 (L.-,`)

Typed by Clover Redfern for Richard Waltz Harper

'

ACKNOWLEDGMENT

I would like to express my deep gratitude to Dr.

Bert Christensen for his help and encouragement in the

preparation of this thesis. Frank Lehmkuhl and Lou

Scerbo also loaned invaluable assistance, particularly

in the performance of elemental analyses, and gathering

of proton magnetic resonance data. I would also like to

thank my wife, Susie, for her encouragement, under-

standing, and help in the preparation of the bibliography.

TABLE OF CONTENTS

INTRODUCTION

DISCUSSION

Page

1

14

EXPERIMENTAL 31 The Methylpyridinium Salt: of 1, 6- dihydro -4- hydroxy -1- methyl -5- nitro- 6- oxopyrimidine [III] 31 1,6- dihydro -4- hydroxy -l- methyl -5 -nit ro- 6 -oxo- pyrimidine [IV] 32 Reaction of 4, 6- dimethoxy -5- nitropyrimidine with pyridine in ethanol 32 1, 6-dihydro-4- methoxy -1 -methyl -5 -nit ro- 6 -oxo- pyrimidine [V] 33 4- Methoxy -l- methyl -5- nitro- 6- oxo -1, 4, 5, 6- tetrahydropyrimidine [VI] 36

BIBLIOGRAPHY 37

LIST OF FIGURES

Figure Page

1. Synthetic route to 4, 6- dimethoxy -5- nitropyrimidine. 15

2. The PMR spectrum of the methylpyridinium salt of 1, 6- dihydro -4- hydroxy -l- methyl -5- nitro -6- oxopyrimidine [III]. 18

3. The PMR spectrum of methylpyridinium iodide. 18

4. The PMR spectrum of 1, 6- dihydro -4- hydroxy -l- methyl-5- nitro -6- oxopyrimidine [IV]. 21

5. The PMR spectrum of 4- methoxy - l - methyl - 5- nitro -6- oxo-1, 4, 5, 6- tetrahydropyrimidine [VI].

6. The IR spectrum of 1, 6- dihydro -4- methoxy -l- methyl -5- nitro -6- oxopyrimidine [V].

7. The IR spectrum of 1, 6- dihydro -4- hydroxy -l- methyl -5- nitro -6- oxopyrimidine [IV].

21

24

24

-

A STUDY OF CERTAIN NOVEL REACTIONS OF 4, 6- DIMETHOXY- 5- NITROPYRIMIDINE

INTRODUCTION

Heterocyclic compounds containing annular nitrogen atoms, like

other amines, readily undergo reaction with a variety of reagents to

form N- methyl derivatives. Several general accounts of work in this

area are available for reference (6, 13, 40, 47).

Methyl halides, and particularly methyl iodide, have been the

most extensively used reagents for N- methylation of nitrogen hetero-

cycles, especially in the formation of quaternary salts. The order of

reactivity of methyl halides in this reaction, which is an example of

the Menshutkin Reaction, is I >Br »» Cl (30). While methyl fluoride has

not been reported to undergo this reaction, methylations with methyl

iodide proceed smoothly under a variety of conditions. Williams, in

1856, reported quaternization of quinoline by simply heating it with

methyl iodide (52). Reactions of this type, with an excess of the

methylating agent as the solvent are generally run in a sealed tube,

due to the volatility of methyl iodide. In some instances, polar apro-

tic solvents have been used, since, a high dielectric constant helps ac-

comodate the charged products, while association with the product,

as through hydrogen bonding, is minimized.

Aromatic nitrogen heterocycles with tautomeric groups in the

2

a- position, generally undergo base catalyzed N- methylation to yield

the reduced N- methyl derivative. However, since these compounds

are ambident in nature, methylation at the exocyclic site is often a

competing side reaction, and in some cases becomes the principal

reaction, Methyl iodide in alcoholic potassium hydroxide has been

reported to convert 2- hydroxypyrimidine to 1, 2- dihydro -l- methyl -2-

oxopyrimidine, plus a small amount of the 2- methoxy derivative (10).

An unusual transannular methylation in the pteridine series was re-

ported by Brown and Jacobsen (7). Heating 2- amino- 4- hydroxypter-

idine with methyl iodide in methanol for 12 hours resulted in conver-

sion to 2- amino -4, 8- dihydro -8- methyl -4- oxopteridine.

OH

NH

The general synthetic utility of various methyl alkyl sulfates,

and aryl sulfonates in N- methylation of nitrogen heterocycles, has

also been well established. Dimethyl sulfate, being less volatile than

methyl iodide, and thus a more convenient reagent in many cases, has

been particularly widely used. A mixture of the 1- and 3- methyl de-

rivatives was obtained from 4- hydroxy -5- phenylpyrimidine on treat-

ment with dimethyl sulfate and alkali (11). Reaction of 4- amino -5-

nitro-2- hydroxypyrimidine with dimethyl sulfate yielded 1, 2- dihydro-

CH3

NIVNH2

O

N

4- amino -l- methyl -5- nitro -2- oxopyrimidine, which is an important

intermediate in purine and pteridine synthesis (20).

>

H3C,NO2

O1\INH 2

Many methylated purines have been identified in recent years in

material from living organisms. Since these compounds are most

likely methylated within the organism, the conditions in the labora-

tory preparations, most of which are carried out in basic medium,

are quite different from those necessarily imposed on in vivo methy-

lation. Jones and Robins, in an attempt to more closely duplicate

physiological conditions, developed a method of methylation under

near neutral conditions, without basic catalysis (21). Dimethylaceta-

mide was chosen as the solvent, as most natural purines display at

least some solubility in this polar solvent, and also because it acts as

a buffer in the presence of dimethyl sulfate, and other methylating

agents. Treatment of xanthine (2, 6- dihydroxypurine) with dimethyl

sulfate in dimethylacetamide at 150 led to 7, 9- dimethylxanthine in

good yield. That later compound had been prepared previously, in

lesser yield, by heating xanthine in a sealed tube with methyl iodide

(5). Rate of reaction is dependent on both the alkylating agent and the

IiNO2 HO

3

HO

H2

4

solvent system, as well as other conditions. The sterically hindered

OH ; CH3 NN N'' -N

)N. i\\T>

HO I

i-:\>

HO I

H CH 3

compound 8- nitroquinoline, is unreactive with methyl iodide, but is

converted to its quaternary salt by dimethyl sulfate (12).

Morley and Simpson attempted to quaternize 4- amino -6-

nitroquinazoline with methyl iodide, but obtained no well defined pro-

duct (35). Experiments with dimethyl sulfate in methanol also failed

to give the desired product. Berg, however, found that fusion with

methyl p- toluenesulfonate at 140°C, or heating with the same reagent

in nitrobenzene at 170 °C, produced the 4- amino -l- methyl -6- nitro-

quinazolinium salt in good yield (1). A vigorous reaction between pi-

peridine and methyl p- toluenesulfonate, has been reported to give the

1-methyl derivative in 74% yield (45). Conversion of 6- aminouracil

to its 3- methyl derivative through the action of methyl p- toluenesul-

fonate and sodium hydroxide in aqueous methanol, has also been ob-

served (50). The unsubstituted methyl benzenesulfonate, with 3-

methylxanthine, in alcoholic potassium hydroxide, produced a 75%

yield of theobromine (38).

Bases which are difficult to quaternize with other reagents,

O

3

> I N / N

5

have been methylated in good yield by fusion with the methyl esters of

o- nitrobenzenesulfonic acid, or 2, 4- dinitrobenzenesulfonic acid.

While 2- methyl -6 -nitro sobenzothiazole is unreactive with dimethyl

sulfate or methyl p- toluenesulfonate, it is converted to the corres-

ponding quaternary salt in 17% yield, by treatment with methyl o-

nitrobenzenesulfonate for two hours at 65°C(25). The quaternary salt

was obtained in 74% yield when the same compound was reacted with

methyl 2, 4- dinitrobenzenesulfonate at ambient temperature for 24

hours. In addition to the above work, Kiprianov and co- workers also

conducted rate studies with a number of methylating agents in various

solvents. They found methyl o- nitrobenzenesulfonate to be six times

faster than dimethyl sulfate, and the methyl 2, 3- and 2, 4- dinitro-

benzenesulfonates to react 60 times as fast as dimethyl sulfate. Ki-

netic studies using 2- methylbenzothiazole substrate, were run in a

toluene solvent. Values for rate constants agreed with a bimolecular

mechanism.

Billman and Cash have reported the use of trimethyl phosphate

in dimethylformamide, for the conversion of 4- nitrophthalidimide to

its N- methyl derivative (3), although this reagent does not seem to

have attracted too much general interest. Trimethyl phosphate gives

the same product as methyl iodide, and like dimethyl sulfate, has the

advantage of being less volatile.

>

O CH

3

6

Diazomethane, which has been used largely for esterification

of carboxylic acids, and generation of carbenes, is also known to pro-

duce N- methyl derivatives of nitrogen heterocycles. The products of

alkylation with diazomethane, however, in some cases differ from

those obtained with other methylating agents. Other methods are pre-

ferable if they give the desired result; since diazomethane, besides

being toxic, is shock sensitive, and consequently prone to explosion.

Diazomethane in ether, reacts with guanine in methanol to yield the

1- methyl derivative, plus the 6- methoxy derivative (16); whereas,

treatment with dimethyl sulfate leads to the expected 7- methyl deri-

vative. Prins, in an attempt to reproduce earlier work (34), found

that 3- hydroxypyridine in an ice water slurry, forms a heterogeneous

system, with diazomethane in ether, which reacted to give about a 10%

yield of the 3- methoxypyridine, and about 30% of 1- methyl -3- hydroxy-

pyridine which was isolated as the picrate (43). The desired 3-

methoxy derivative was produced by the action of diazomethane in t-

butyl alcohol, on a homogeneous system formed with an ethereal solu-

tion of 3- hydroxypyridine, in 68% yield. Among the pyrimidine

7

reactions, it has been reported that 6- methyl -4- hydroxypyrimidine,

upon treatment with diazomethane yielded the 3- methyl derivative,

plus a little of the 4- methoxy derivative (32).

Lewis acids such as boron trifluoride, and fluoroboric acid,

have been found to catalyze the reaction of diazomethane with amines

and alcohols. However, these Lewis acids also catalyze an unwanted

side reaction involving polymerization of diazomethane (33), which is

minimized through the use of an excess of the amine. Five parts of

piperidine, with 1 part boron trifluoride, when reacted in ether with

diazomethane, gave 1- methyl -piperidine in 30% yield, and 60% poly-

mer. In contrast to this, reaction of 23 parts of piperidine and 1

part of boron trifluoride, with diazomethane, yielded 39% 1- methyl-

piperidine, with no isolable polymer (36).

Reductive alkylation (14) employing formaldehyde and a hydro-

gen source has also been used in the preparation of N- methyl deriva-

tives of nitrogen heterocycles. Piperazine, on treatment with form-

aldehyde, in the presence of zinc and hydrochloric acid, was con-

verted to 1, 4- dimethylpiperazine in 88% yield (15). The Eschweiler-

Clarke Modification of the Leuckart Reaction uses formic acid as the

hydrogen source, in conjunction with formaldehyde. Clarke and co-

workers reported the methylation of piperidine under these conditions,

to produce the 1- methyl derivative in over 80% yield (9). Castrillion

has reported cyclization in the course of methylation of a ring

8

substituted ß- phenylethyl amine, with formaldehyde and formic acid,

to yield a substituted 1, 2, 3, 4- tetrahydroisoquinoline as the only pro-

duct (8). This occurrence is a special case of the Pictet -Spengler

Synthesis.

H3

H3CO NH2 i

CH H

3 CO

3 CH 3

Irreversible rearrangement of lactim ethers to their lactam

forms, catalyzed thermally or chemically, also provides a useful

synthetic route to N- methylated nitrogen heterocycles. Knorr found

that certain 2- and 4- alkoxyquinolines formed unstable addition pro-

ducts with methyl iodide, which on heating, decomposed to yield the

N- alkyloxoquinolines (27). On heating with methyl iodide, 2- ethoxy-

quinoline was found to be converted to : 1- ethyl- 2- oxoquinoline.. Re-

arrangements of a similar nature were noted by Lieben and Haitinger

in the hydroxypyridines (31). Hilbert and Johnson established the

synthetic utility of this rearrangement in the pyrimidine series (18).

They found that 2, 4- dimethoxypyrimidine, on heating at 220°C for four

hours, gave a quantitative yield of 1, 3- dimethyluracil. A mixture of

five grams of 2, 4- dimethoxypyrimidine and ten grams of methyl io-

dine, yielded 1- methyl -4- methoxy -2- oxopyrimidine quantitatively, on

standing at room temperature for two to three hours. In the same

CH 3

9

reaction mixture, diluted with benzene, discoloration was observed,

and reaction was not complete for several weeks. Using five grams

of the pyrimidine, and two drops of methyl iodide in a sealed tube,

the reaction was complete in two days at ambient temperature. Addi-

tion of pyridine in catalytic or stoichiometric amounts, has been re-

ported to improve the yield of this reaction, in cases where it was

previously less than stoichiometric (44). This rearrangement has

also found great synthetic utility in the Hilbert -Johnson Synthesis of

nucleosides (42). Among the purine reactions, Bergman and Heim -

hold have noted that 2, 6- dimethoxy -7- methylpurine, on heating, or

treatment with methyl iodide, was converted to 1, 3, 6- trimethyl -2, 6-

dioxopurine (2).

H _CO

>

CH 3

An intermolecular rearrangement similar to the aforementioned

intramolecular lactim -lactam rearrangement, has been patented (28).

This reaction employes the 0- methyl ether of caprolactim as the

methylating agent. Uric acid (2, 6, 8- trihydroxypurine), on heating

for 18 hours with 100 parts of caprolactim 0- methyl ether, was con-

verted to tetramethyl uric acid. A number of other aromatic nitrogen

OCH3CH3

\ /\ N

O CH3 HC

3 \N

o%\NN

10

heterocycles were found to react in a similar fashion.

Methanol in the presence of Raney Nickel, in a bomb at elevated

temperature, has also been shown to give N- methylation. Plante and

Clapp (41) have reported that 2, 2, 5, 5- tetramethyl piperazine, with

methanol and Raney Nickel in a bomb at 200°C, was converted to 1, 2,

2, 4, 5, 5- hexamethylpiperazine in 46% yield. Under the same condi-

tions 2, 3, 4, 5- tetrahydro -2, 2, 5, 5- tetramethylpyrazine, and 5- amino-

3-aza-2, 2, 5- trimethyl- l- hexanol, were both also converted to the

above product. Cyclization of this type of compound in the presence

of Raney Nickel had been previously observed (26).

CH H3C OH H C

H3C \/NH

CH CH3

H

CH3 CH3

CH3

Methyl salicylate has been found to be synthetically useful in

the quaternization of aromatic nitrogen heterocycles (23). Pyridine

was converted to the corresponding quaternary salt by reaction with

methyl salicylate at 120 -30°C for five hours. The investigators hy-

pothesized that this unexpected reaction proceeded because of the

stabilization of the resulting salicylate ion through hydrogen bonding

with the o- hydroxyl group. The reaction was unsuccessful using

methyl .benzoate as the methylating agent, lending credibility to the

I 3

11

above hypothesis. Methyl cyanoacetate was found to methylate pyri-

dine, indicating that acid strength of the carboxylic acid involved is

an important consideration. Subsequent work has shown that this re-

action proceeds with a variety of methyl esters of carboxylic acids,

and confirmed the idea that the yield of quaternary salt produced is

proportional to the acid strength of the carboxylic acid from which

the ester is derived (24).

In some cases, methyl trichloroacetate has also been found to

act as a methylating agent (39). The trifluoro- and tribromoacetates

were also found to convert triethyl amine to the corresponding N-

methyl triethylammonium acetate; however, in the case of the tri-

fluoroacetic acid methyl ester, amide formation was a competing

side reaction. These reagents appear to be of limited synthetic val-

ue, in view of the fact that with primary and secondary amines, only

amides and urethanes are formed (22).

A novel N- methylation reaction of 2, 5- dimethylpyrazine, in-

volving chloroacetic acid as the methyl precursor, has been reported

(17). This reagent generally reacts with tertiary amines to form the

betaine, as in the case of pyridine, which is a stable product. With

N

-

\ I \ % --,,'"

CH2CO2

N_

12

2, 5- dimethylpyrazine, however, bromo- or chloroacetic acid reacts

with the evolution of carbon dioxide to form the N- methyl derivative.

Unsubstituted pyrazine, and quinoxaline, react with some evolution

of carbon dioxide, but the tarry product isolated was not the desired

N- methyl quaternary salt. This isolated instance is possible due to

mesomeric participation of the second ring nitrogen, which being in

the para position, can help support the incipient negative charge.

The current investigation arose from work concerning an ab-

normal reaction of 4, 6- dimethoxy -5- nitropyrimidine [I], with methyl -

hydrazine, which was first observed by Krakov, in normal butyl al-

cohol (29). Instead of the expected 4, 6-di(1- methylhydrazino) -5-

nitropyrimidine, a few milligrams of an unknown crystalline solid,

and a large amount of starting material were isolated. Krakov hy-

pothesized that the unknown product was 3- methyl- (1- methylhydra-

zino)-3H-v- triazolo [4, 5 -d] pyrimidine 1- oxide.

I

OCH I 3 H39TNH2

NO2 CH3NHNH2 NO2

3 2) N 1

N

CH3

H3CNNH2 o

N

\ NOCH3

CH3NHNH2

CH 3

II

NNH2

3

T

1.\T7---.

N \

3

13

Stahl (49) investigated this reaction in several solvents, and

found that the unknown material was produced in good yield using re-

fluxing pyridine as the solvent, in about one hour. Subsequently,

Stahl identified the unknown product as 4- hydrazino -6- hydroxypyrim-

idine [II], which he synthesized in an unambiguous manner. It was

also found in the course of this investigation, that I reacted with re-

fluxing pyridine to produce another compound [III], to which we as-

signed the following structure on the basis of its carbon and hydrogen

analysis, proton magnetic resonance (PMR) spectrum, and infrared

(IR) spectrum:

O H3C\rNO2 \ I NO

III

CH3 I+ N

Upon dissolving III in fresh pyridine, and treating with methylhydra-

zine, II was not formed.

The subject of the current investigation is the unusual reaction

resulting in III, and the possible general synthetic utility of I as an

N- methylating agent.

14

DISCUSSION

In order to obtain more III for purposes of confirmation of the

structure suggested by Stahl, it was first necessary to synthesize

some 4, 6- dimethoxy -5- nitropyrimidine. This was done according to

the scheme depicted in Figure 1. Malondiamide was made to undergo

base catalyzed cyclization with ethyl formate and sodium ethoxide in

ethanol, according to the method of Hull (19), to yield 4, 6- dihydroxy --

pyrimidine. The later compound was nitrated in the 5- position by the

procedure of Boon, Jones and Ramage (4), using glacial acetic acid

and red fuming nitric acid. Chlorination of the 4, 6- dihydroxy -5-

nitropyrimidine was accomplished employing phosphorus oxychloride,

as reported by the same workers (4). However, N, N- diethylaniline

was used in place of N, N- dimethylaniline as the catalyst, a modifica-

tion which was reported by Krakov (29) to result in increased yield.

It may be noted here for later reference, that addition of the N, N-

diethylanìline to the 4, 6- dihydroxy -5- nitropyrimidine previous to

addition of the phosphorus oxychloride, was reported by Stahl (49), to

result in a highly exothermic polymerization reaction, accompanied

in large scale preparations by uncontrollable foaming. Treatment of

4, 6- dichloro -5- nitropyrimidine with a cold solution of sodium meth -

oxide in methanol, resulted in conversion to 4, 6- dimethoxy -5- nitro-

pyrimidine [I], as described by Rose and Brown (46).

15

Cl

HCOOCH2CH3

NaOCH2CH3

OH

CH3COOH (glacial)

HNO 3

OH

O_

'N OH

POC13 C

N, N-diethylaniline

NaOCH3

Figure 1. Synthetic route to 4, 6- dimethoxy -5- nitropyrimidine.

O

{

H 2N CHZ

H 0 2N i

OCHS

N NO2

N ~OCH3

N \\

OH

II\

N

16

Upon refluxing I in excess pyridine it was converted to a com-

pound having the properties of III in about 25% yield, melting point

155 -7°C. Evaporation of the pyridine yielded only a red oil, which

was found by Stahl to be converted to II, upon dissolving in fresh

pyridine followed by treatment with methylhydrazine. No attempt was

made to elucidate the structure of the oil.

The PMR spectrum of III showed that the two equivalent methoxy

methyl groups of I, absorbing at 4.126, had split into a three proton

singlet of 3. 426 and another at 4. 426. The spectrum of III was ob-

served in deuterium oxide, in which it is very soluble, while the spec-

trum of I, which is very insoluble in water, was taken in deutero-

chloroform. The difference in shift between the two solvents is not

significant. This change was accompanied by disappearance of the

strong absorption band in the 1125 cm -1 region of the IR spectrum,

which had been attributed to the methoxy groups in I. An N- methyl

group is suggested by the band at 3. 425 in the PMR spectrum, as

would result from a lactim -lactam rearrangement analogous to that

reported by Hilbert and Johnson in the 2, 4- dihydroxypyrimidines

(18), which has been shown to be catalyzed by pyridine (44). Absorp-

tion bands due to the N- methyl groups of 1, 3- dimethyluracil, shown

in spectrum number 461 of the Varian NMR spectra catalogues, ap-

pear at 3. 306 and 3. 436 (51). Further evidence is provided by the

appearance of an absorption band at about 1650 cm -1 in the IR

17

spectrum of III, which is indicative of a conjugated amide carbonyl

group (37), such as is present in the proposed structure. Methylation

of the pyridine annular nitrogen to form the methylpyridinium portion

of III, could be responsible for the three proton singlet at 4. 426.

Synthesis of methylpyridinium iodide, by mixing pyridine with methyl

iodide at ambient temperature, provided further evidence inasmuch

as a three proton singlet for the N- methyl group of the methylpyri-

dinium iodide appeared at 4. 466. The absorption pattern of methyl

pyridinium iodide in deuterium oxide between 8 and 96 were also found

to be identical to that observed for III, except for apparent enhance-

ment by one proton in the spectrum of III, in the region of 8..106, due

to the overlap with the 2- proton of the pyrimidine ring. This methyl -

ation of the solvent may be analogous to the previously cited work of

Kametani and co- workers (23, 24), involving N- methylation of nitro-

gen heterocycles with the methyl esters of certain organic acids as

the methyl precursors. Later in the discussion this possibility will

be dealt with in greater depth.

The salt like nature of the hypothetical III suggested that exam-

ination of the hydrolysis products might provide further support for

this structure. Hydrolysis of III in dilute hydrochloric acid yielded a

tan crystalline compound IV, which melted with decomposition in the

range 265 -9°C. , to which was assigned the structure 1, 6- dihydro -4-

hydroxy- l -methyl- nitro -6- oxopyrimidine. Carbon and hydrogen

18

l0

ppm (6 )

Figure 2. The PMR spectrum of the methylpyridinium salt of 1, 6- dihydro- 4- hydroxy -1 -methyl -5- nitro -6- oxopyrimidine [III].

0

10 9

Figure 3.

8 7 6 5 4 3 2 1

ppm(6)

The PMR spectrum of methylpyridinium iodide.

9 8 7 6 5 4 3 2 1

19

analysis supported an empirical formula of C5H5N304, which is con-

sistent with the proposed structure.

HC1 H3CN

V I

H20

CH 3

Ñ+ Cl

IV

In deuterium oxide -deuterated sodium hydroxide, the PMR

spectrum of IV appeared as two singlets at 8. 476, and 2. 836, present

in a ratio of one to three. While the integration data is consistent

with the proposed structure of IV, the N- methyl absorption would be

expected to occur at lower, rather than higher field with respect to

the N- methyl group on the pyrimidine portion of III, due to the induc-

tive effect of the negative charge in III. The possibility of base cata-

lyzed ring cleavage, to which 4, 6- dihydroxypyrimidines appear to be

particularly disposed, was considered as a likely explanation for

H3C NO2

OH

NaOD D20

H C\DOG NO2 "ND 3 ND

I

OH

the observed spectrum. Subsequently, deuterated sodium hydroxide,

which was thought to be essential to increase the solubility of IV in

III +

tOI

2

O

2

IxOI

N

20

deuterium oxide, was found to be unnecessary. In deuterium oxide,

the PMR spectrum of N showed a singlet of 9. 076, and another at

3.626, again integrating one to three, which is more consistent with

the proposed structure.

An attempt was made to examine the PMR spectrum of 4, 6-

dihydroxy-5- nitropyrimidine, in deuterium oxide, as a possible model

for the absorption of the 2- proton of N, which seemed further down -

field than might be expected. However, no absorption was observed,

indicating that the proton had exchanged with the solvent deuterons

rather readily. This suggests that the 2- proton is indeed reasonably

acidic, and should appear considerably shifted downfield.

With the structures of III and IV reasonably established, exten-

sion of the reaction to other solvent systems was attempted. The ab-

normal reaction of I with methylhydrazine was first observed by Kra-

kov in normal butyl alcohol solvent, so the possibility of reaction be-

tween I and that solvent was explored. Refluxing I in normal butyl

alcohol for three hours resulted in recovery of 94% of the starting

material as identified by its melting point and IR spectrum. After

refluxing for 36 hours, 84% of the starting material was recovered.

Upon evaporation of the solvent from the later reaction under reduced

pressure, a small amount of another white crystalline material was

isolated, melting point greater than 300°C.; although, it began to de-

compose about 230oC. Examination of the infrared spectrum of this

21

10 9 8 7 6 5 4 4

3 2 1 0

ppm(6 )

Figure 4. The PMR spectrum of 1, 6- dihydro -4- hydroxy -1 - methyl -5- nitro -6- oxopyrimidine [IV].

I n

10 9 8 7 6 5 4 3 2 1 0

ppm(6)

Figure 5. The PMR spectrum of 4- methoxy -1- methyl -4- nitro- 6- oxo -1, 4, 5, 6- tetrahydropyrimidine [VI].

f t I t- f I e I

22

material showed it to be different from any of the other compounds

involved in this study, and possibly a mixture. Since insufficient ma-

terial was isolated for analysis, and this solvent showed little pro-

mise, the reaction was not pursued further. This did show, however,

that no reaction between I and the solvent is necessary in the forma-

tion of II from I and methylhydrazine in the original reaction described

by Stahl.

To ascertain if the basicity of the solvent might be an important

factor, the reaction of I with quinoline was examined. A flask fitted

with a reflux condensor, and magnetic stirrer bar, containing 1 g

of I and 50 ml of quinoline, was placed in an oil bath maintained at

135 -40°C. Although the mixture turned dark very quickly, the heat-

ing was continued for 45 minutes. The dark solution was then refrig-

erated for several hours, at which time a trace of black precipitate,

melting point greater than 300°C. , was isolated. Due to the difficulty

involved in removal of this high boiling solvent, the reaction was run

in the same fashion, using only 25 ml of quinoline. Still no significant

amount of precipitation occurred.

An effort to run the reaction in some solvent inert to the reac-

tion conditions was made at this time, as a possible way of minimiz-

ing the isolation problem. Absolute ethanol was chosen as a reason-

ably volatile, polar solvent, which would not be acidic enough to react

with the expected salt, nor basic enough to react with I. A mixture

23

of 1 g of I, plus 0.45 ml, an equimolar quantity of pyridine, in 25

ml of absolute ethanol was refluxed with stirring for three hours.

Cooling the reaction mixture in a refrigerator for several hours, then

filtering with suction, resulted in recovery of 95% of the starting ma-

terial as identified by its melting point and IR spectrum.

The apparent unsuitability of solvents other than excess sub-

strate, led to experiments with a limiting amount of substrate. Reactions

of I with pyridine, in a one to two molar ratio was first attempted.

Heating 1 g of I,, and 0.85 ml of pyridine in an oil bath maintained

at 110 -5oC. , for 30 minutes, resulted in a dark yellow oil, which

crystallized upon cooling and agitation. The IR spectrum of the pro-

duct was similar to that of III, but indicated contamination, as was

confirmed by its melting range of 128 -34°C. Treatment with water,

and suction filtration of the resulting mixture, yielded a white, water

insoluble, crystalline solid [V] which decomposed slowly starting

about 220°C. , but melted above 300°C. The infrared spectrum of V

showed bands at about 1412, 1497, 1600, and 1635 cm -1, which indi-

cate that the pyrimidine nucleus is still intact (48). A strong band at

1682 cm -1 supports the presence of an amide type carbonyl function

(37). While being somewhat reminicent of the spectrum of IV, the

decided difference in the fingerprint region of the spectrum led to the

hypothesis that V is the 1, 6- dihydro -4- methoxy -5- nitro -6- oxopyrimi-

dine, since 1, 3- dimethylation of this compound is prohibited by its

WAVENUMBER CM -t 2000 100400 000 1200 CO

I 1000

0o-

90

60

70

60

50

40-

30- 20

M-

25

Figure

4000

00 -

90

60

5 6 7

WAVELENGTH IN MICRONS 10

6. The IR spectrum of 1, 6- dihydro -4- methoxy -1- methyl -5- nítro-6- oxopyrimidine [V].

WAVENUMBER CM -1 3000 2500 2000 100 A00300 200 1100 000 400 @00 700 550 625

70

60-

50- . 40

30- ¡ 20

b 0

25 3 4 5 6 7 8

WAVELENGTH IN MICRONS

Figure 7. The IR spectrum of 1, 6- dihydro -4- hydroxy -1- methyl -5- nitro- 6- oxopyrimidine [IV].

0`

3

550 524

25

valence requirements.

The solubility behavior of V is such that its PMRspectrum could

only be taken in deuterium oxide - deuterated sodium hydroxide, which

as was previously noted, reacts with this type of compound, resulting

in ring cleavage. The PMR spectrum showed two singlets, at 2. 935

and 8. 575, present in a ratio of six to one. Equivalence of the two

methyl groups suggests that the ring cleavage reaction was followed

by an intermolecular reaction to yield the N, N- dimethyl compound.

This follows from the fact that the spectrum is essentially identical

to that of IV in deuterium oxide -deuterated sodium hydroxide, except

that the N- methyl peak, at 2. 935, was doubled in size.

H3C NO2

O OCH3

H N OD OD

D O N D N GH3 2 H G

O

D

H3CCH3

OD

Elemental analysis revealed a hydrogen and nitrogen content

consistent with the formula C6H7N3O4; whereas , the carbon was low

and variable, indicating poor combustion of the compound.

In view of the low carbon analysis, it seemed advisable to at-

tempt reduction of the compound to its amino derivative, which should

combust more easily, and be more soluble for PMRanalysis, to pro-

vide a confirmation for the structure of V. A Parr low pressure

\ NO

26

hydrogenation apparatus was used, employing 10% palladium on char-

coal as the catalyst, and methanol as the solvent. Removal of the

solvent under reduced pressure, revealed a red oil, which gave a

yellow crystalline solid VI when dissolved in water. Recrystalliza-

tion from water resulted in isolation of VI as a white crystalline ma-

terial, m. p. 138 -9°C. , which was supposed to be the 5- amino -1, 6-

dihydro-l- methyl -6- oxopyrimidine, in relatively poor yield. Ele-

mental analysis, however, indicated an empirical formula of

C6H9N3O4, supporting reduction of a double bond rather than reduc-

tion of the nitro group, to give 4- methoxy -l- methyl -5- nitro- 6 -oxo-

1, 4, 5, 6- tetrahydropyrimidine.

H3C H

OCH 3

NO2

VI

This later structure is also supported by the PMR spectrum of

VI. In deuterochloroform, the spectrum appeared as a broad singlet

at 3. 27; a mutliplet centered at about 4. 806, and a singlet at 5. 806,

integrating six to two to one. The 3. 276 band can be assigned as due

to the overlap of the N- and 0- methyl groups, the 4. 806 band to the

4- and 5- protons, which split each other, and the 5. 806 band to the

2- proton. If the other double bond had been reduced, the methoxy

H

27

methyl group should appear further downfield, and the 4. 806 band

should appear as a singlet.

With the structure of V reasonably established, the other pro-

ducts of the reaction were examined. The aqueous solution from

which V was taken, when evaporated to dryness under reduced pres-

sure, yielded again yellow oily crystals. Extraction of these crys-

tals with cold pyridine removed the oil, yielding O. 56 g of dry

yellow crystals. This later material was identified as pure III, in

40% yield, by melting point, infrared spectrum, and by hydrolysis in

dilute hydrochloric acid to IV, as identified by its melting point and

infrared spectrum.

Evaporation of the pyridine used in the extraction noted above,

revealed a dark, thick oil, which did not crystallize upon heating in

vacuo for several hours, then cooling. This oil was dissolved in

fresh pyridine, and treated with methylhydrazine, to determine wheth-

er it would lead to 4- hydrazino- 6- hydroxypyrímidine, as did the oil

isolated from the reaction of I with excess pyridine. This reaction

led to no isolable product except oil, upon evaporation under reduced

pressure.

Reaction of I with pyridine was then tried on an equimolar basis,

and subjected to the same work -up procedure as the previous reaction.

This led to isolation of V in 48% yield, III in 32% yield, and a small

amount of oil which did not lead to II when treated with methylhydrazine.

28

Thus, the reaction of I with pyridine, is found to change course

as the amount of pyridine present is varied. When excess pyridine is

present as a solvent, no V is isolated; and the amount of V isolated

is found to increase as the amount of pyridine decreases. This could

be at least in part, due to the viscosity of the fusion type reaction

medium involved in reaction of I with a small amount of pyridine. In

the very viscous oil formed, the ions formed by initial attack of pyri-

dine on I, would be very slow to diffuse apart, providing ample time

for the methyl group to be transferred from the pyridinium ion back

to the ring nitrogen of the pyrimidine to form V. It was also found

that upon refluxing V in fresh pyridine, no isolable III was formed.

The unreactivity of V towards pyridine would then explain why V ac-

cumulates more rapidly as the amount of solvent is decreased.

Experiments involving reaction of I with other nitrogen hetero-

cycles on an equimolar basis were then performed. Heating I with

one equivalent of piperidine, led to isolation of V in 19. 6% yield, plus

a thick, dark oil, which was unaffected by acid. Upon refluxing this

oil with fresh piperidine, and methylhydrazine, no II could be isolat-

ed.

In the course of this reaction, it was found that heating the re-

action mixture too rapidly, or too hot, results in a violently exother-

mic reaction, which sprays material out the top of the condensor.

This leads to a black, insoluble material as the only isolable product.

29

Apparent decomposition of a similar nature was also noted in other

equimolar reactions, but was minimized by heating the bath slowly,

to a temperature not over about 120oC. There seems to be an analogy

between this observation, and the exothermic, foaming reaction be-

tween 2, 4- dihydroxy -5- nitropyrimidine, and N, N- diethylaniline ob-

served by Stahl, as previously noted.

Quinoline was then reacted with I in an equimolar ratio. The

resulting dark oil was heated in vacuo to produce a black, foamy

mass, which broke up into an apparently crystalline material, which

could be recrystallized from water to give V in about 15% yield. The

black material, recovered from water, and dissolved in fresh quino-

line, was heated with methylhydrazine at about 120 -5°C. No II could

be isolated from this reaction, or from similar reactions involving

the black decomposition products from equimolar experiments.

Reaction of I with 4- methylpyrimidine, run and worked up in the

same manner as the reaction with quinoline, yielded about the same

results. A 10% yield of V was isolated, plus the black, decomposition

products previously noted. No further investigation of the later ma-

terial was undertaken.

Thus, 4, 6- dimethoxy -5- nitropyrimidine does not seem promis-

ing as a general reagent for the N- methylation of nitrogen heterocy-

cles. Reactions involving I, with excess substrate as the solvent lead

in some cases to problems in separating the products from the

30

reaction mixture, and in any case, do not lead to high yields of iso-

lable N- methylated product. Side reactions, apparently involving de-

composition, or polymerization, become major reactions when I is

fused with an equimolar amount of substrate, and the only isolable

product is the rearranged pyrimidine, 1, 6- dihydro -4- methoxy -l-

methyl -5 -nit r o -6 - oxopyrimidine.

31

EXPERIMENTAL

All melting points were taken on a Fischer -Johns melting point

apparatus, and are uncorrected. Infrared spectra were obtained

through the use of a Beckman Model IR -8 spectrometer, using sodi-

um chloride plates. These spectra were calibrated with the 6. 2461.1

band of polystyrene. The infrared samples were run as mulls in nu-

jol oil (nujol itself absorbs strongly at about 2920 cm -1, and

2860 cm -1, and less intensly at about 1460 cm -1, and 1375 cm -1).

A Varian A -60 spectrometer produced the proton magnetic resonance

spectra. Tetramethylsilane was used as an internal standard for

samples run in deuterochloroform. An external standard composed

of 10% tetramethylsilane in deuterochloroform was used for samples

dissolved in deuterium oxide, unless otherwise noted.

The Methylpyridinium Salt of 1, 6- dihydro -4- hydroxy -1 - methyl -5- nitro -6- oxopyrimidine [III]

In a 100 ml round - bottom flask, were placed 1. 027 g (0.0056

mole) of 4, 6- dimethoxy -5- nitropyrimidine, prepared by the method

of Rose and Brown (46) and 50 ml of anhydrous pyridine. The flask

was fitted with a reflux condensor, and heated in an oil bath main-

tained at 135 -40oC. , for about 40 minutes, with stirring. On cooling

overnight in a refrigerator, the salt precipitated as a brown crystal-

line solid. The mixture was filtered with suction, and the solid

32

washed with fresh pyridine, and dried irr vacuo at about 80°C. , for 24

hours, yielding 0. 379 g (25.8 %) of the light brown salt, m. p. 155 -7°C.

Anal. Calc'd for C11H12N4O4: C, 49. 99; H, 4. 59.

Found: C, 49.81; H, 4.56.

1, 6- dihydro -4- hydroxy -1 - methyl -5- nitro -6- oxopyrimidine [IV]

To 0.379 g of the methylpyridinium salt of 1, 6- dihydro -4-

hydroxy -1- methyl -5- nitro -6- oxopyrimidine, was added five ml of di-

lute hydrochloric acid, whereupon, a tan precipitate formed immedi-

ately. The mixture was refrigerated for several hours, then filtered

with suction. Recrystallization of the solid from water, yielded

0. 118 g (43. 3 %) of 1, 6- dihydro -4- hydroxy -1- methyl -5- nitro -6-

oxopyrimidine, m. p. (dec.) 265 -9 °C.

Anal. Calc'd for C5H5N304: C, 35. 09; H, 2. 95.

Found: C, 34.83; H, 2.89.

Reaction of 4, 6- dimethoxy-5- nitropyrimidine with pyridine in ethanol

A mixture of 1. 035 g (0. 0.056 mole) of 4, 6- dimethoxy -5-

nitropyrimidine, 0. 45 ml (0. 0056 mole) of pyridine, and 25 ml of ab-

solute ethanol, in a 50 ml round- bottom flask, fitted with reflux con -

densor, and magnetic stirrer bar, was heated at reflux for three

hours with stirring. Cooling the mixture, and filtering with suction,

.

33

resulted in recovery of Q. 978 g (94.5%) ; of the starting material,

as identified by its IR spectrum and melting point.

1, 6- dihydro- 4- methoxy -1 -methyl -5- nitro -6- oxopyrimidine [V]

A. In a ten ml round -bottom flask, were placed 0. 978 g

(0. 0053 mole) of I, and two equivalents of pyridine, 0. 834 g (0.85 ml).

The flask was fitted with a reflux condensor, and placed in an oil bath,

which was heated to 110 -5oC. , and maintained at that temperature

for about 30 minutes. A dark viscous oil resulted, which upon cool-

ing and agitation, formed yellow, hygroscopic crystals. When these

crystals were extracted with water, a relatively water insoluble cry-

stalline compound was obtained. This later compound recrystallized

from water as 0.194 g (19. 5 %) of pale yellow crystals of V, which,

dried in vacuo, melted above 300°C. , although gradual decomposition

was observed above about 230oC.

Anal. Calc'd for C6H7N3O4: C, 38. 92; H, 3. 82; N, 22. 70.

Found: C, 37.65; H, 3.85; N, 22.67.

The water with which the crystals were extracted above, was

evaporated under reduced pressure to again yield oily yellow crystals.

These crystals were extracted with pyridine, which removed the oil,

to yield 0. 556 g of III, m. p. 155 -7°C. This compound was further

identified by its IR spectrum, and by hydrolysis to IV, which was re-

cognized by its melting point and IR spectrum. .

.

34

Under reduced pressure, the pyridine used in the extraction

above was evaporated to yield a yellow oil. This oil was dissolved in

15 ml of fresh pyridine, and combined with one ml of methylhydrazine.

After standing for three days no isolable 4- hydrazino -6- hyroxypyrim-

idine had formed.

B. A mixture of 1. 047 g (0. 0056 mole), of I, and 0. 54 ml

(0. 0056 mole), of pyridine, equimolar amounts, was reacted, and

worked up as described above. A 48% yield of V, 0. 500 g, was

obtained. The salt III was isolated in 32% yield, 0. 474 g. The oil

which was isolated, again did not lead to II.

C. In a ten ml round -bottom flask, fitted with reflux condensor,

were placed 1. 004 g of I (0. ,005.4 mole), and 0. 54 ml (0. 0054 mole)

of piperidine. The flask was placed in an oil bath, which was heated

slowly to 110 °C. , and maintained at 110 -5 °C. , for about 45 minutes,

at which time a thick, dark oil had formed. On cooling, the oil solid-

ified, but did not crystallize. With addition of about five ml of water,

a white precipitate formed. Recrystallization from water yielded

0..197 g (19 6 %)of V, m. p, greater than 30Q °C.

Evaporation under reduced pressure, as previously described,of

the aqueous solution failed to give any solid material yielding only the

oil. This oil was unaffected by acid, and upon dissolving in 25 ml of

fresh piperidine, and treatment with methylhydrazine at room

,

35

temperature for two days, yielded no isolable II.

D. 4, 6- dimethoxy -5- nitropyrimidine, 1. 026 g (0. 0055 mole)

and 0. 66 ml (0. 0055 mole) of quinoline, in a ten ml round- bottom

flask equipped with reflux condensor, were placed in an oil bath,

which was heated to 108oC. , then maintained at 108 -12oC. , for about

one hour. The resultant oil was heated in vacuo, in an oil bath main-

tained at 80 -90oC. , for two days. A black foam formed, which broke

up into fine crystals. Treatment of this black material with water

resulted in 0.137 g (13. 4%); of V, after recrystallization from wa-

ter.

E. 4- methylpyrimidine, 0.. 488 g (0. 0.054 mole), 'and 1. Q12

g of 4, 6- dimethoxy -5- nitropyrimidine (0. 0054 mole), were reacted

and worked up as described in the previous procedure. After recry-

stallization from water, 0. 128 g (12.7%) of V, m. p.. greater than

300°C.

The product isolated from each of the previous procedures,

showed identical infrared absorption spectra, and elemental analysis.

The carbon was found to be low and slightly variable presumably due

to incomplete combustion.

36

4- Methoxy -l- methyl -5- nitro- 6- oxo -1, 4, 5, 6- tetrahydropyrimidine[VI]

In a Parr low pressure hydrogenation apparatus, were placed

1.124 g (0 0061 moles) of V, suspended in 150 ml of anhydrous

methanol, and 0. 057 g of 10% palladium on charcoal catalyst, un-

der an initial pressure of 15 psi. . The flask was then agitated for

about ten hours, at which time all of the starting material had gone

into solution, and the pressure had dropped indicating a reaction had

taken place. The pressure drop was not significant enough to calcu-

late moles of hydrogen absorbed. The solution was then heated to an

incipient boil, and the catalyst removed, and washed with fresh

methanol. The combined methanol solution was evaporated under re-

duced pressure, yielding a reddish -brown oil. Addition of a small

amount of water to the oil resulted in a white, crystalline material.

Recrystallization from water gave VI, 0.. 16.9 g (14. 8%), m. p« 1.38 -

9oC.

Anal. Calc'd for C6H9N3O4: C, 38. 50; H, 4. 86; N, 22. 45.

Found: C, 38.66; H, 4.94; N, 22.18.

37

BIBLIOGRAPHY

1. Berg, S. S. A search for new trypanocides. Part VI. Amidino- phenyldiazoamino- quinolinium and -quinazolinium salts. Journal of the Chemical Society, 1961, p. 4041 -4045.

2. Bergmen, Ernest and Heinz Heimhold. Rearrangement of meth - oxy- pyrimidines and -purines. Journal of the Chemical Society, 1935, p. 955 -957.

3. Billman, John H. and Vincent R. Cash. 4- Nitrophthalidimide. III. Methylation with trimethyl phosphate and hydroxymethyla- tion with formaldehyde. Proceedings of the Indiana Academy of Science 63:108 -109. 1953.

4. Boon, W. R. , W. G. M. Jones and G. R. Ramage. Pteridines. Part I. An unambiguous synthesis of 7:8- dihydro -6- hydroxy- pteridines. Journal of the Chemical Society, 1951, p. 96 -102.

5. Bredereck, Hellmut and Gottfried Kupsch. Systhesen in der Pur - inreihe, VIII. "Desoxyharnsäuren ". Ihre Konstitution als Xanthiniumbetaine und neue Synthesen. Chemische Berichte 92:566 -582. 1959.

6. Brown, D. J. The pyrimidines. New York, Interscience, 1962. 774 p.

7. Brown, D. J. and N. W. Jacobsen. Unique transannular methyla- tion of 2- amino -4- hydroxypteridine. Tetrahedron Letters, 1960, no. 25, p. 17 -20.

8. Castrillion, Jose A. Cactus alkaloids. II. Condensation of mescaline with formaldehyde by the Eschweiler- Clarke reaction. Journal of the American Chemical Society 74 :558 -559. 1952.

9. Clarke, H. T. , H. B. Gillespie and S. Z. Weisshaus. The action of formaldehyde on amines and amino acids. Journal of the American Chemical Society 55:4571 -4587. 1933.

10. Curd, F. S. H. and D. N. Richardson. The synthesis of trypano- cides. Part IV. The attempted preparation of 4- amino -6 -(6- amino -1: 2 - dimethylpy r imidinium- amino) -1: 2 - dimethylquino - linium di- iodide. Journal of the Chemical Society, 1955, p. 1853- 1858.

38

11. Davies, W.H. and H.A. Piggott. A novel pyrimidine synthesis. Part I. 4- Amino -5- phenylpyrimidine. Journal of the Chemical Society, 1945, p. 347 -351.

12. Decker, Herman. Ueber einige Ammoniumverbindung. (11. Mitteilung) Ueber 8- Nitrochinolinjodmethylat. Berichte der Deutschen Chemischen Gesellschaft 36 :261. 1903.

13. Duffin, G. F. The quaternization of heterocyclic compounds. Advances in Heterocyclic Chemistry 3 :1 -56. 1964.

14. Emerson, William S. The preparation of amines by reductive alkylation. In: Organic reactions, ed. by Roger Adams et al. New York, Wiley, 1948. p. 174 -255.

15. Forsee, W. T. , Jr. and C. B. Pollard. Derivatives of pipera - zine. VI. Alkylation by means of aldehydes. Journal of the American Chemical Society 57 :1788 -1789. 1935.

16. Friedman, Orrie M., G.N. Mahapatra and Robert Stevenson. The methylation of ribonucleo side s by diazomethane. Biochimica et Biophysica Acta 68:144 -146. 1963.

17. Hart, E. V. and P. E. Spoerri. The decarboxylation of pyrazin- ium betaine salts. Journal of the American Chemical Society 77 :5898 -5899. 1955.

18. Hilbert, Guido E. and Treat B. Johnson. Researches on pyrimi- dines. CXV. Alkylation of the nitrogen of the pyrimidine cycle . . by application of a new technique involving molecular rearrange- ments. Journal of the American Chemical Society 52 :2001 -2007. 1930.

19. Hull, R. A new synthesis of 4:6- dihydroxypyrimidines. Journal of the Chemical Society, 1951, p. 2214.

20. Johns, Carl O. Researches on purines. On 2-oxy-1- methyl- purine. The Journal of Biological Chemistry 11:73 -79. 1912.

21. Jones Jesse W. and Roland K. Robins. Potential purine antag- onists. XXX. Purine betaines and related derivatives prepared by direct methylation of simple purines. Journal of the American Chemical Society 84:1914 -1918. 1962.

.

39

22. Joullie, Madeleine M. and Allan R. Day. Effect of structure on reactivity. IX. A study of the aminolysis of esters of trichloro- and trifluoroacetic acids. Journal of the American Chemical Society 76:2990 -2993. 1954.

23. Kametani, Tetsuji et al. A novel methylation of tertiary amines with methyl salicylate. Tetrahedron Letters, 1965, no. 23, p. 1817 -1822.

24. Kametani, Tetsuji et al. Synthesis of heterocyclic compounds. CLIV. Novel methylation. 3. Methylation of tertiary amines, such as pyridine and isoquinoline with alkyl carboxylates. The Journal of Heterocyclic Chemistry 3 :129 -136. 1966.

25. Kiprianov, A. I. and A. I. Tolmachev. Preparation of quaternary salts of weak organic bases. Zhurnal Obshchei Khimii 27:142- 150. 1957. (Abstracted in Chemical Abstracts 51: 12912f. 1957)

26. Kitchen, Leland J. and C. B. Pollard. Derivatives of piperazine. XXI. Synthesis of piperazine and c- substituted piperazines. Journal of the American Society 69:854 -855. 1947.

27. Knorr, Ludwig. Lieber das Verhalten cues Antipyrins gegen Halogealkyle. Justus Liebig's Annalen der Chemie 293:1-41. 1896.

28. Konz, Wilhelm. N- alkylation. German patent no. 944, 312.. June 14, 1956. (Abstracted in Chemical Abstracts 52: 16379d. 1958)

29. Krakov, Mark Harry. The cyclization reactions of certain 5- amino -4- chloro -6- hydrazinopyrimidines with phosgene. Ph. D. thesis. Corvallis, Oregon State University, 1962. 86 numb. leaves.

30. Larsen, Robert P. and Charles A. Kraus. The rates of reaction between alkyl halides and tertiary amines as determined conduc- tometrically. Proceedings of the National Academy of Sciences 40:70- 76. 1954.

31. Lieben, A. and L. Haitinger. Untersuchen über Chelidonsaüre. Monatshefte für Chemie und vervandte Teile anderer Wissen- schaften 6 :279 -328. 1885.

40

32. Marshall, J. R. and James J. Walker. An experimental study of some potentially tautomeric 2- and 4(6)- substituted pyrimi- dines. Journal of the Chemical Society, 1951, p. 1004 -1017.

33. Meerwein, Hans. Uber die katalytische Zersetzung des Diazo- methans und das Polymethylen. Angewandte Chemie 60:78. 1948.

34. Meyer, Hans. Uber die Einwirkung von Diazomethan auf Pyri- done und Oxypyridincarbonsäuren. Monatshefte für Chemie und verwante Teile anderer Wissenschaften 26 :1311 -1329. 1905.

35. Morley, J. S. and J. C. E. Simpson. The chemistry of simple heterocyclic systems. Part I. Reactions of 6- and 7- nitro -4- hydroxyquinoline and their derivatives. Journal of the Chemical Society, 1948, p. 360 -366.

36. Müller, Eugen, Helmut Huber -Emden and Wolfgang Rundel. Untersuchungen an Diazomethanen. XI. N- methylierung von Aminen mit Diazomethan. Justus Liebig's Annalen der Chemie 623 :34 -46. 1959.

37. Nakanishi, Koji. Infrared absorption spectroscopy. San Fran- cisco, Holden -Day, 1962. 233 p.

38. Nestrov, V. M. and I.E. Chubova. Methylation of 3- methylxan- thine with methyl benzenesulfonate. Khimiko- Farmatsevticheskii Zhurnal 1:42 -43. 1967. (Abstracted in Chemical Abstracts 67: 3071t. 1967)

39. Pierce, Arleen C. and Madeleine M. Joullie. Reactions of tri - halogenated esters with triethyl amine and anions. The Journal of Organic Chemistry 27:3968 -3973. 1962.

40. Pilyugin, G. T. and B. M. Gutsulyak. Advances in the prepara- tion, investigation, and use of quinolinium compounds. Russian Chemical Reviews 32: 167 -188. 1963. (Translated from Uspekhi Khimii)

41. Plante, Laurence T. and Leallyn B. Clapp. Alkylations with alcohol and raney nickel catalyst. The Journal of Organic Chem- istry 21:86 -87. 1956.

42. Pliml, Jiri and Miroslav Pyrstas. The Hilbert- Johnson reaction of 2, 4- dialkoxypyrimidines with halogenoses. Advances in Het- erocyclic Chemistry 8 :115 -142. 1967.

41

43. Prins, D. A. Methylation of 3- hydroxypyridine. Recueil des Travaux Chimiques des Pays -Bas 76:58 -64. 1957.

44. Rabinowitz, Joseph L. and Samuel Gurin. Some pyrimidine de- rivatives. Journal of the American Chemical Society 75:5758- 5759. 1953.

45. Reynold, D. D. and W. O. Kenyon. Preparation and reactions of sulfonic esters. V. Synthesis of cyclic tertiary amines. Jour- nal of the American Chemical Society 72:1597-1598. 1950.

46. Rose, F. L. and D. J. Brown. An anomalous reaction of 4,6- dichloro-5-nitropyrimidine.. Journal of the Chemical Society, 1956, p. 1953 -1956.

47. Shaw, Elliot N. Quaternary pyridinium compounds. In: Pyri- dine and its derivatives, ed. by E. Klingsberg. Part II. New York, Interscience, 1964. p. 1 -95.

48. Short, L. N. and H. W. Thompson. Infrared spectra of deriva- tives of pyrimidine. Journal of the Chemical Society, 1952, p. 168 -187.

49. Stahl, Quade Russell. Unpublished research on hydrazinopyri- midines. Corvallis, Oregon, Oregon State University, Dept. of Chemistry, 1967.

50. Ukai, Teiji, Yuzuru Yamamoto and Shoichi Kanetomo. Xanthine derivatives... I. Methylation of 4- aminouracil. Journal of the Pharmaceutical Society of Japan 74:674 -677. 1954. (Abstracted in Chemical Abstracts 48: 10743f. 1954)

51. Varian Associates. High resolution NMR spectra catalog. Vol. 2. Palo Alto, Calif. , 1963. Various paging.

52. Williams, G. Chinolin und Homologe desselben. Jahresbericht über die Fortschritte der Chemie und verwandter Thiele anderer Wissenschaften, 1856, p. 532 -537.